Everybody "knows" that neural networks need more than a single layer ofnonlinear units to compute interesting functions. We show that this is false if one employs winner-take-all as nonlinear unit: - Any boolean function can be computed by a single k-winner-takeall unitapplied to weighted sums of the input variables.

In order to to compare learning algorithms, experimental results reported in the machine learning litterature often use statistical tests of significance. Unfortunately,most of these tests do not take into account the variability due to the choice of training set. We perform a theoretical investigation of the variance of the cross-validation estimate of the generalization errorthat takes into account the variability due to the choice of training sets. This allows us to propose two new ways to estimate this variance. We show, via simulations, that these new statistics perform well relative to the statistics considered by Dietterich (Dietterich, 1998). 1 Introduction When applying a learning algorithm (or comparing several algorithms), one is typically interested in estimating its generalization error. Its point estimation is rather trivial through cross-validation. Providing a variance estimate of that estimation, so that hypothesis testing and/orconfidence intervals are possible, is more difficult, especially, as pointed out in (Hinton et aI., 1995), if one wants to take into account the variability due to the choice of the training sets (Breiman, 1996). A notable effort in that direction is Dietterich's work (Dietterich, 1998).Careful investigation of the variance to be estimated allows us to provide new variance estimates, which tum out to perform well. Let us first layout the framework in which we shall work.

The algorithm builds on the basic principles of decomposition proposed by Osuna et.

The problem that we address in this paper is how a mobile robot can plan in order to arrive at its goal with minimum uncertainty. Traditional motion planning algorithms oftenassume that a mobile robot can track its position reliably, however, in real world situations, reliable localization may not always be feasible. Partially Observable Markov Decision Processes (POMDPs) provide one way to maximize the certainty of reaching the goal state, but at the cost of computational intractability for large state spaces. The method we propose explicitly models the uncertainty of the robot's position as a state variable, and generates trajectories through the augmented pose-uncertainty space. By minimizing the positional uncertainty at the goal, the robot reduces the likelihood it becomes lost. We demonstrate experimentally that coastal navigation reduces the uncertainty at the goal, especially with degraded localization.

We develop a hierarchical generative model to study cue combination. Themodel maps a global shape parameter to local cuespecific parameters,which in tum generate an intensity image. Inferring shape from images is achieved by inverting this model. Inference produces a probability distribution at each level; using distributions rather than a single value of underlying variables at each stage preserves information about the validity of each local cue for the given image. This allows the model, unlike standard combination models, to adaptively weight each cue based on general cuereliability and specific image context.

We find that a necessary requirement for effective associative memorylearning is that the efficacies of the incoming synapses should be uncorrelated. This requirement is difficult to achieve in a robust manner by Hebbian synaptic learning, since it depends on network level information. Effective learning can yet be obtained by a neuronal process that maintains a zero sum of the incoming synapticefficacies. This normalization drastically improves the memory capacity of associative networks, from an essentially bounded capacity to one that linearly scales with the network's size. It also enables the effective storage of patterns with heterogeneous coding levels in a single network.

We describe a class of probabilistic models that we call credibility networks. Using parse trees as internal representations of images, credibility networks are able to perform segmentation and recognition simultaneously,removing the need for ad hoc segmentation heuristics. Promising results in the problem of segmenting handwritten digitswere obtained.

I consider a topographic projection between two neuronal layers with different densitiesof neurons. Given the number of output neurons connected toeach input neuron (divergence or fan-out) and the number of input neurons synapsing on each output neuron (convergence or fan-in) I determine the widths of axonal and dendritic arbors which minimize the total volume ofaxons and dendrites. My analytical results can be summarized qualitativelyin the following rule: neurons of the sparser layer should have arbors wider than those of the denser layer. This agrees with the anatomical data from retinal and cerebellar neurons whose morphology andconnectivity are known. The rule may be used to infer connectivity ofneurons from their morphology.

The project pursued in this paper is to develop from first information-geometric principles a general method for learning the similarity between text documents.

I describe a silicon network consisting of a group of excitatory neurons anda global inhibitory neuron. The output of the inhibitory neuron is normalized with respect to the input strengths.